Electroluminescence (EL) is the emission of light from a phosphor due to the application of an electric field. Electroluminescent devices have utility as lamps and displays. Currently, electroluminescent devices are used in flat panel display systems, involving either pre-defined character shapes or individually addressable pixels in a rectangular matrix.
Pioneering work in electroluminescence was done at GTE Sylvania. An AC voltage was applied to powder or dispersion type EL devices in which a light emitting phosphor powder was imbedded in an organic binder deposited on a glass substrate and covered with a transparent electrode. These powder or dispersion type EL devices are generally characterized by low brightness and other problems which have prevented widespread use.
Thin film electroluminescent (TFEL) devices were developed in the 1950's. The basic structure of an AC thin layer EL laminate is well known, see for example Tornqist, R.O "Thin-Film Electroluminescent Displays" Society for Information Display, 1989, International Symposium Seminar Lecture Notes, and U.S. Pat. No. 4,857,802 to Fuyama et al. A phosphor layer is sandwiched between a pair of electrodes and separated from the electrodes by respective insulating/dielectric layers. Most commonly, the phosphor material is ZnS with Mn included as an activator (dopant). The ZnS:Mn TFEL is yellow emitting. Other colour phosphors have been developed.
The layers of conventional TFEL laminates are deposited on a substrate, usually glass. Deposition of the layers is done sequentially by known thin film techniques, for example electron beam (EB) vacuum evaporation or sputtering and, more recently, by atomic layer epitaxy (ALE). The thickness of the entire TFEL laminate is only in the order of one or two microns.
To separate and electrically insulate the phosphor layer from the electrodes, various insulating/dielectric materials are known and used, as discussed in more detail. hereinafter.
Each of the two electrodes differ, depending on whether it is at the "rear" or the "front" (viewing) side of the device. A reflective metal, such as aluminum is typically used for the rear electrode. A relatively thin optically transmissive layer of indium tin oxide (ITO) is typically employed as the front electrode. In lamp applications, both electrodes take the form of continuous layers, thereby subjecting the entire phosphor layer between the electrodes to the electric field. In a typical display application, the front and rear electrodes are suitably patterned with electrically conductive address lines defining row and column electrodes. Pixels are defined where the row and column electrodes overlay. Various electronic display drivers are well known which address individual pixels by energizing one row electrode and one column electrode at a time.
While simple in concept, the development of thin film electroluminescent devices has met with many practical difficulties. A first difficulty arises from the fact that the devices are formed from individual laminate layers deposited by thin film techniques which are time consuming and costly techniques. A very small defect in any particular layer can cause a failure. Secondly, these thin-film devices are typically operated at relatively high voltages, eg. 300-450 volts peak to peak. In fact, these voltages are such that the phosphor layer is operated beyond its dielectric breakdown voltage, causing it to conduct. The thin-film dielectric layers on either side of the phosphor layer are required to limit or prevent conduction between the electrodes. The application of the large electric fields can cause electrical breakdown between the electrodes, resulting in failure of the device.
The present invention is particularly directed to the insulating/dielectric layers of electroluminescent devices and the prevention of electrical discharges across the phosphor layer. A requirement for successful operation of an electroluminescent device is that the electrodes (address lines) be electrically isolated from the phosphor layer. This function is provided by the insulating/dielectric layers. Typically, insulating/dielectric layers are provided on either side of the phosphor layer and are constructed from alumina, yttria, silica, silicon nitride or other dielectric materials. During operation of the device, electrons from the interface between the insulating layer and the phosphor layer are accelerated by the electric field as they pass through the phosphor layer, and collide with the dopant atoms in the phosphor layer, emitting light as a result of the collision process. In a conventional TFEL device, to ensure that the electric field strength across the phosphor is sufficiently high, the thickness of the dielectric layers is usually kept less than or comparable to that of the phosphor layer. If the dielectric layers are too thick a large portion of the voltage applied between the address lines is across the dielectric layers rather than across the phosphor layer.
It is important that the dielectric material be compatible with the phosphor layer. By "compatible", as used in this specification and in the claims, is meant that, firstly, it provides a good injectivity interface, i.e. a source of "hot" electrons at the phosphor interface which can be promoted or tunnelled into the phosphor conduction band to initiate conduction and light emission in the phosphor layer on application of an electric field. Secondly, within the meaning of compatible, the dielectric material must be chemically stable so that it does not react with adjacent layers, that is the phosphor or the electrodes.
In a typical TFEL, in order to achieve sufficient luminosity, the applied voltage is very near that at which electrical breakdown of the dielectric occurs. Thus, the manufacturing control over the thickness and quality of the dielectric and phosphor layers must be stringently controlled to prevent electrical breakdown. This requirement in-turn makes it difficult to achieve high manufacturing yields.
A typical TFEL structure is constructed from the front (viewing) side to the rear. The thin layers are sequentially deposited on a suitable substrate. Glass substrates are utilized to provide transparency. The transparent, front electrode (ITO address lines) is deposited on the glass substrate by sputtering to a thickness of about 0.2 microns. The subsequent dielectric--phosphor--dielectric layers are then usually deposited by sputtering or evaporation. The thickness of the phosphor layer is typically about 0.5 microns. The dielectric layers are typically about 0.4 microns thick. The phosphor layer is usually annealed after deposition at about 450.degree. C. to improve efficiency. The rear electrode is then added, typically in the form of aluminum address lines with a thickness of 0.1 microns. The finished TFEL laminate is encapsulated in order to protect it from external humidity. Epoxy laminated cover glass or silicon oil encapsulation are used. In that the initial substrate used for deposition is typically glass, the materials and deposition techniques employed in TFEL laminate construction cannot demand high temperature processing.
The high electric field strength used to operate a TFEL device puts heavy requirements on the dielectric layers. High dielectric strengths are required to avoid electrical breakdown. Dielectrics with high dielectric constants are preferred in order to provide luminosity at the lowest possible driving voltage.. However, efforts to utilize high dielectric constant materials have not provided satisfactory results.
To lower the driving voltage of TFEL elements insulating layers have been constructed from higher dielectric constant materials, for instance SrTiO.sub.3, PbTiO.sub.3, and BaTa.sub.2 O.sub.3, as reported in U.S. Pat. No. 4,857,802 issued to Fuyama et al. However, these materials have not performed well, exhibiting low dielectric breakdown strengths. In U.S. Pat. No. 4,857,802, a dielectric layer is formed from a perovskite crystal structure by controlled thin film deposition techniques to achieve an increased (111) plane orientation. The patent reports higher dielectric strengths (above about 8.0.times.10.sup.5 -- about 1.0.times.10.sup.6 V/cm) with a dielectric layer having a thickness of about 0.5 microns using SrTiO.sub.3, PbTiO.sub.3 and BaTiO.sub.3, all of which have high dielectric constants and a perovskite crystal structure. This device still has the disadvantage of requiring complex and difficult to control thin film deposition techniques for the dielectric layer.
Efforts have also been made to develop TFEL devices using a thick ceramic insulator layer and a thin film electroluminescent layer, see Miyata, T. et al., SID 91 Digest, pp 70-73 and 286-289. The device is built up from a BaTiO.sub.3 ceramic sheet. The sheet is formed by molding fine BaTiO.sub.3 powder into disks (20 mm diameter) by conventional cold-press methods. The disks are sintered in air at 1300.degree. C. then ground and polished into sheets with a thickness of about 0.2 min. The emitting layer is deposited onto the sheet in a thin film using chemical vapour deposition or RF magnetron sputtering. Suitable electrode layers are then deposited by thin film techniques on either side of the structure. While this device exhibits certain desirable characteristics, it is not feasible to manufacture a commercial TFEL device from a solid ceramic sheet. Grinding and polishing a larger ceramic sheet to a consistent thickness of 0.2 mm is not practical economically.
It is also known in the art to use multiple insulating/dielectric layers on each side of the phosphor layer. For instance, U.S. Pat. No. 4,897,319 to Sun discloses a TFEL with an EL phosphor layer sandwiched between a pair of insulator stacks, in which one or both of the insulator stacks includes a first layer of silicon oxynitride (SiON) and a second thicker layer of barium tantalate (BTO). The first, SiON layer provides high resistivity while the second., BTO layer has a higher dielectric constant. Overall, the structure is stated to produce a higher luminance of the phosphor layer at conventional voltages. However, the insulating layers are deposited by RF sputtering, which has the disadvantages of thin film techniques described hereinabove.
There is a need for a TFEL device having higher luminosity and lower operating voltage than conventional TFEL devices, while still being feasible to construct. It is necessary to achieve this with a dielectric layer which has a dielectric strength that is above the electric field strength needed to drive the device.
Fabricating electrode patterns in transparent conductor materials such as indium tin oxide often involves extensive and expensive masking, photolithographic and chemical etching processes. Lasers have been proposed for scribing such transparent conductor materials. Generally carbon dioxide, argon and YAG lasers are used. Such lasers produce light in the visible and infrared ranges of the electromagnetic spectrum (generally greater than 400 nm). However, there are problems in using such long wavelength light to scribe electrode patterns, particularly when the transparent conductor material is deposited on another transparent layer. In conventional TFEL displays, the transparent electrode material, typically indium tin oxide (ITO), is deposited on the transparent display glass (substrate) prior to depositing the remaining layers of the EL laminate. In an insulator or a semiconducting material, light with a wavelength longer than that corresponding to the energy of the electronic band gap in the material is not strongly absorbed. For optically transparent materials, the wavelength corresponding to the band gap is shorter than that for visible light. Therefore, transparent electrode materials show poor absorption of laser light due to both the long wavelength of the light and the thinness of the layer, making it difficult to utilize laser energy to directly ablate the electrode address lines.
U.S. Pat. No. 4,292,092, to Hanak and U.S. Pat. No. 4,667,058, to Catalano et al., disclose processes to pattern a transparent electrode pattern deposited on another transparent layer in a solar battery. The patents teach patterning the electrode using a pulsed YAG laser, which produces light with a wavelength too long to be significantly absorbed in any of the transparent layers. To compensate for the low absorption, a laser with high peak power is used to thermally vaporize the transparent electrode. A neodymium YAG laser is operated at 4-5 W with a pulse rate of 36 KHz at a scanning rate of 20 cm/sec. The examples of the patent disclose scribing an ITO layer deposited on glass in this manner. However, the scribed lines are described as having incompletely removed the ITO and, in places, as having melted the glass to a depth of a few hundred angstroms. The residual ITO must thereafter be removed by a subsequent etching step.
Other approaches to forming electrode patterns in transparent electrode materials involve using an excimer laser, which produces light of shorter wavelength, in the ultraviolet region of the electromagnetic spectrum. At this wavelength, the laser energy can be absorbed by the transparent electrode material. Lasers of this nature are suggested to form conductive patterns for liquid crystal displays (U.S. Pat. No. 4,980,366, to Imatou et al and U.S. Pat. No. 4,927,493, to Yamazaki et al.), photovoltaic batteries (U.S. Pat. No. 4,783,421, to Carlson et al. and U.S. Pat. No. 4,854,974, to Yamazaki et al.) imaging sensors (U.S. Pat. No. 5,043,567, to Sakama et al.), and integrated circuits (U.S. Pat. No. 5,109,149, to Leung). WO 90/0970, published Aug. 23, 1990, to Autodisplay A/S, discloses a process for scribing an electrode dot matrix pattern in a transparent conductor on a transparent substrate with an excimer laser.
While excimer lasers produce light which has a wavelength short enough to be absorbed by the transparent electrode such that the electrode may be patterned by direct ablation, such lasers are relatively expensive and the scribing process must be carefully controlled to avoid melting or ablating the underlying display glass. Furthermore, such processes may lead to excessive or incomplete ablation of the transparent electrode material. For instance in WO 90/0970 there is an indication that, in the event of partial removal 0f the material to be ablated, remaining portions may be removed by chemicals or plasma etching.
Another problem encountered in scribing transparent electrode materials on a transparent substrate is addressed in U.S. Pat. No. 4,937,129, to Yamazaki. To avoid diffusion or cross contamination between the layers, diffusion barrier layers are provided at the interface.
Other patents have taught surface treatments of the transparent electrode material to enhance absorption of the laser light. For instance, U.S. Pat. No. 4,909,895, to Cusano, teaches oxidizing the metallic film surface to make it less reflective of the laser light. U.S. Pat. No. 4,568,409, to Caplan, teaches coating the transparent layer to be ablated with a dye to selectively absorb laser light where ablation is desired.
Control circuitry to drive an EL display has been developed. Basically, the circuitry converts serial video data into parallel data to apply a voltage to the rows and columns of the display. State of the art row and column driver components (chips) are available.
Asymmetric and symmetric drive techniques are used with EL displays. In an asymmetric drive method, the EL panel is provided with drive pulses by applying a negative subthreshold voltage to one row at a time. During each row scan time, a positive voltage pulse is applied to the selected columns (i.e. those that should illuminate) and zero voltage is applied to the nonselected columns (i.e. those that should not illuminate). At the intersection of selected columns and rows, a voltage equal to the sum of the subthreshold row voltage and the positive pulse voltage on the column is applied across the pixel, causing light emission. After all rows of the panel have been addressed, a positive polarity refresh pulse is applied to all of the rows simultaneously, and all columns are held at 0 V.
In a symmetrical drive scheme, the refresh pulse is eliminated. Instead, a similar set of drive pulses that are of the opposite polarity are applied to the panel. To maintain the panel in operation, the rows are scanned with pulses of alternating polarity on even and odd frames. The alternating polarity produces a net zero charge on all display pixels.
State of the art high voltage driver components (chips) are available for both asymmetric and symmetric drive techniques.
Alternate driving circuits and components for EL displays are known or are in development, see for example K. Shoji et al, Bidirectional Push-Pull Symmetric Driving Method of TFEL Display, Springer Proceedings in Physics, Vol. 38, 1989, 324: and Sutton S. et al, Recent Developments and Trends in Thin-Film Electroluminescent Display Drivers, Springer Proceedings in Physics, Vol. 38, 1989, 318; and Bolger et al, A Second Generation Chip Set for Driving EL Panels, SID, 1985, 229.
The above driving schemes are termed multiplexed (passive) matrix addressing schemes. Theoretically, other types of driving Schemes, such as active matrix addressing schemes, could be used with EL displays. However, these are not yet developed. Such alternate driving schemes should be considered to be within the meaning of the phrase voltage driving circuitry as used in this application.
In conventional EL displays, one method to connect the column and row address lines to the driver circuit is to compress a polymeric strip containing very many closely spaced metal sheets between rows of contacts connected to the display address lines and rows of contacts connected to the driver components of the driver circuit, which is constructed on a separate circuit board (see U.S. Pat. No. 4,508,990, to Essinger). The polymeric strip is a layered elastomeric element (LEE), known by such tradenames as STAX and ZEBRA. The LEE is composed of alternating layers of conductive and nonconductive elastomeric materials. The polymeric strip avoids the need to laboriously connect hundreds of individual wires using solder or welded connections to the contacts. However, this interconnection technology is unreliable, and does not function well at high temperatures, which can cause the polymeric material to creep.
Another method that is commonly used to connect column and row address lines to the driver circuit for liquid crystal displays (LCDs) is being considered for electroluminescent displays, namely chip-on-glass (COG) technology. The driver components (chips) to which the address lines must be connected are mounted around the periphery of the display. In the case of LCDs, the address lines, which are evaporated on the rear side of the display glass, are extended from the active region of the display so that they end in contact pads that are arranged in a pattern so that the chips can be wire bonded thereto. Wire bonding entails mounting the chips on the display glass and then individually welding fine gold wires to the output pads on the chip and to the corresponding contact pads on the address lines.
The advantage of COG technology is that the number of contacts between the display glass and the driver circuit are substantially reduced, since by far the largest number of contacts are between the driver chips and the address lines. There are typically only about 20 to 30 connections between the driver chips and the rest of the driving circuit as opposed to up to 2000 connections to the address lines.
One major disadvantage of the COG technology is that difficulty is experienced in wire bonding the driver chips to connect them to the thin film pads on the address lines, resulting in poor manufacturing yields. Another disadvantage is that space is required around the perimeter of the display to mount the driver chips, thus increasing the bulkiness of the displays and eliminating any possibility of joining several display modules in an array to form a larger display.
Through hole technology for direct circuit connections is widely known in the semiconductor art (see for example U.S. Pat. No. 3,641,390, Nakamura). U.S. Pat. No. 4,710,395, to Young et al, describes methods and apparatus for through hole substrate printing with regulated vacuum. However, through hole printing has not, to the inventors' knowledge, been successfully applied to EL displays.
U.S. Pat. No. 3,504,214 to Lake et al describes a segmented storage type of EL device in which pixels are turned on with light to make a photoconductive layer next to the phosphor layer become electrically conductive. Complex through hole conductors are described. The patent indicates that ordinary through hole connections do not work with high resolution TFEL displays because the conductive material might react with the phosphor, thereby degrading the performance of the display.